capacitance vs emi

what's worse for messing up your waveform for automotive diagnostics, uncompensated capacitance, or electromagnetic interference? to use back probes, i can either use a bnc to banana jack adapter and 2 banana leads, or i can use the bnc coax cable that at the end of it breaks out into 2 banana plugs.  the latter i'm guessing uses the ground conductor as shielding for the signal wire, but probably at the expense of increased capacitance.

There is no one-fits-all answer, it all depends.

The frequency and intensity of the interference play a mayor part in this equation,as well as the kind and magnitude of the measured signal is.
Another facor is the shielding characteristic of the coax. Then, even a standard banana lead will have some capacitance.

From a practical standpoint:
As you seem to have both options just try which one works better for you when a real problem shows up with one of them.

Depending on your scope and familiarity with it: up to certain point active filtering does help in aquiring "cleaner" signal.

In modern cars, except the ignition signals, you do not have so much electromagnetic perturbations.

So if you are considering to mesure very very low voltage variations or signals at high frequencies near perturbations frequencies, it should be a major issue.

I have both shielded and non shielded probes withous problems. Just be carreful for current probe (as amp clamps) that may be affected.